System and method for supporting common transaction identifier (XID) optimization based on resource manager (RM) instance awareness in a transactional environment

Abstract

A system and method can support transaction processing in a transactional environment. A coordinator for a global transaction operates to propagate a common transaction identifier and information for a resource manager instance to one or more participants of the global transaction in the transactional environment. The coordinator allows said one or more participants, which share resource manager instance with the coordinator, to use the common transaction identifier, and can process the global transaction for said one or more participants that share the resource manager instance using one transaction branch.

Description

CLAIM OF PRIORITY

This application claims priority on U.S. Provisional Patent Application No. 61/985,135, entitled “COMMON TRANSACTION IDENTIFIER (XID) OPTIMIZATION BASED ON DATABASE INSTANCE AWARENESS,” by inventors Xugang Shen and Todd Little, filed Apr. 28, 2014, which application is herein incorporated by reference.

CROSS REFERENCE TO RELATED APPLICATIONS

This application is related to the following patent applications, each of which is hereby incorporated by reference in its entirety:

U.S. patent application titled “SYSTEM AND METHOD FOR SUPPORTING RESOURCE MANAGER (RM) INSTANCE AWARENESS IN A TRANSACTIONAL ENVIRONMENT”, application Ser. No. 14/587,474, filed Dec. 31, 2014;

U.S. patent application titled “SYSTEM AND METHOD FOR SUPPORTING TRANSACTION AFFINITY BASED ON RESOURCE MANAGER (RM) INSTANCE AWARENESS IN A TRANSACTIONAL ENVIRONMENT”, application Ser. No. 14/587,476, filed Dec. 31, 2014; and

U.S. patent application titled “SYSTEM AND METHOD FOR ADAPTIVELY INTEGRATING A DATABASE STATE NOTIFICATION SERVICE WITH A DISTRIBUTED TRANSACTIONAL MIDDLEWARE MACHINE”, application Ser. No. 14/603,108, filed Dec. 31, 2014.

COPYRIGHT NOTICE

A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.

FIELD OF INVENTION

The present invention is generally related to computer systems and software, and is particularly related to a transactional system.

BACKGROUND

A transactional middleware system, or transaction oriented middleware, includes enterprise application servers that can process various transactions within an organization. With the developments in new technologies such as high performance network and multiprocessor computers, there is a need to further improve the performance of transactional middleware. These are the generally areas that embodiments of the invention are intended to address.

SUMMARY

Described herein are systems and methods that can support transaction processing in a transactional environment. A coordinator for a global transaction operates to propagate a common transaction identifier and information for a resource manager instance to one or more participants of the global transaction in the transactional environment. The coordinator allows said one or more participants, which share the resource manager instance with the coordinator, to use the common transaction identifier, and the coordinator can process the global transaction for said one or more participants that share the resource manager instance using one transaction branch.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 shows an illustration of supporting resource manager (RM) instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 2 shows an illustration of maintaining various state tables in a transactional environment, in accordance with an embodiment of the invention.

FIG. 3 shows an illustration of supporting a server table (ST) in a transactional environment, in accordance with an embodiment of the invention.

FIG. 4 shows an illustration of updating the instance information in a transactional environment, in accordance with an embodiment of the invention.

FIG. 5 shows an illustration of supporting a transaction process with various check points in a transactional environment, in accordance with an embodiment of the invention.

FIG. 6 illustrates an exemplary flow chart for supporting resource manager (RM) instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 7 shows an illustration of supporting transaction affinity based on instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 8 shows an illustration of supporting the transaction affinity routing in a transactional environment, in accordance with an embodiment of the invention.

FIG. 9 shows an illustration of sending a message with an affinity context in a transactional environment, in accordance with an embodiment of the invention.

FIG. 10 shows an illustration of receiving a message with affinity context in a transactional environment, in accordance with an embodiment of the invention.

FIG. 11 shows an illustration of supporting affinity routing within a client context in a transactional environment, in accordance with an embodiment of the invention.

FIG. 12 shows an illustration of propagating an affinity context across different domains in a transactional environment, in accordance with an embodiment of the invention.

FIG. 13 shows an illustration of propagating an affinity context to an application server in a transactional environment, in accordance with an embodiment of the invention.

FIG. 14 illustrates an exemplary flow chart for supporting transaction affinity based on instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 15 shows an illustration of processing a global transaction in a transactional environment using different transaction identifiers (XI Ds), in accordance with an embodiment of the invention.

FIG. 16 shows an illustration of processing a global transaction in a transactional environment using a common transaction identifier (XID), in accordance with an embodiment of the invention.

FIG. 17 shows an illustration of supporting a one-phase commit (1PC) processing model based on database instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 18 shows an illustration of processing a global transaction based on database instance awareness in a transactional environment, in accordance with an embodiment of the invention.

FIG. 19 shows an illustration of processing a global transaction across multiple domains in a transactional environment using a common XID, in accordance with an embodiment of the invention.

FIG. 20 illustrates an exemplary flow chart for processing a global transaction in a transactional environment using a common XID, in accordance with an embodiment of the invention.

DETAILED DESCRIPTION

The invention is illustrated, by way of example and not by way of limitation, in the figures of the accompanying drawings in which like references indicate similar elements. It should be noted that references to “an” or “one” or “some” embodiment(s) in this disclosure are not necessarily to the same embodiment, and such references mean at least one.

The description of the invention as following uses the ORACLE® TUXEDO transactional middleware environment (hereinafter TUXEDO) as an example for a transactional middleware machine environment. It will be apparent to those skilled in the art that other types of transactional middleware machine environments can be used without limitation.

Described herein are systems and methods that can support transaction processing in a transactional environment, such as a transactional middleware machine environment.

Transactional Middleware Machine Environment

In accordance with an embodiment of the invention, the system comprises a combination of high performance hardware, e.g. 64-bit processor technology, high performance large memory, and redundant InfiniBand and Ethernet networking, together with an application server or middleware environment, such as ORACLE® WEBLOGIC unified application server software suite, (hereinafter WEBLOGIC) to provide a complete JAVA® EE application server complex which includes a massively parallel in-memory grid, that can be provisioned quickly, and can scale on demand. In accordance with an embodiment, the system can be deployed as a full, half, or quarter rack, or other configuration, that provides an application server grid, storage area network, and InfiniBand (IB) network. The middleware machine software can provide application server, middleware and other functionality such as, for example, WEBLOGIC, JROCKIT JAVA® virtual machine software or HOTSPOT JAVA® virtual machine software (Hotspot JVM), ORACLE® LINUX® operating system software or SOLARIS® operating system software, and Or ORACLE® VM. In accordance with an embodiment, the system can include a plurality of compute nodes, IB switch gateway, and storage nodes or units, communicating with one another via an IB network. When implemented as a rack configuration, unused portions of the rack can be left empty or occupied by fillers.

In accordance with an embodiment of the invention, the system provides an easy-to-deploy solution for hosting middleware or application server software, such as the ORACLE® Middleware SW suite, or WEBLOGIC. As described herein, in accordance with an embodiment the system is a “grid in a box” that comprises one or more servers, storage units, an IB fabric for storage networking, and all the other components required to host a middleware application. Significant performance can be delivered for all types of middleware applications by leveraging a massively parallel grid architecture using, e.g. ORACLE® Real Application Clusters database software and ORACLE® EXALOGIC computing appliance Open storage. The system delivers improved performance with linear I/O scalability, is simple to use and manage, and delivers mission-critical availability and reliability.

In accordance with an embodiment of the invention, a transactional middleware system, such as the ORACLE® TUXEDO system, can take advantage of fast machines with multiple processors, such as an ORACLE® EXALOGIC middleware machine, and a high performance network connection, such as an IB network. Additionally, the ORACLE® TUXEDO system can take advantage of a clustered database, such as the ORACLE® Real Application Clusters (RAC) Enterprise database, which is a clustered database with shared cache architecture and can be a component of a cloud architecture. The ORACLE® RAC can overcome the limitations of traditional shared-nothing and shared-disk approaches to provide highly scalable and available database solutions for business applications.

In accordance with an embodiment of the invention, ORACLE® TUXEDO system provides a set of software modules that enables the construction, execution, and administration of high performance, distributed business applications and has been used as transactional middleware by a number of multi-tier application development tools. TUXEDO is a middleware platform that can be used to manage distributed transaction processing in distributed computing environments. It is a proven platform for unlocking enterprise legacy applications and extending them to a services oriented architecture, while delivering unlimited scalability and standards-based interoperability.

Additionally, the ORACLE® TUXEDO system can comply with the Open Group's X/Open standards, including the support of the XA standard for two-phase commit (2PC) processing, the X/Open ATMI API, and the X/Open Portability Guide (XPG) standards for language internationalization. The transactional application server can be referred to as an XA server, when it uses the XA standard. For example, each TUXEDO application server that belongs to a TUXEDO group can be configured using an OPENINFO property. All XA servers in a TUXEDO group can use the OPENINFO property to establish the connection to a resource manager (RM).

Instance Awareness

FIG. 1 shows an illustration of supporting resource manager (RM) instance awareness in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 1, a transactional system 101 in a transactional environment 100 can support transaction processing using one or more resource manager (RM) instances associated with a data source 102, such as a database.

In accordance with an embodiment of the invention, the transactional system 101 can be aware of the RM instance information 104 in the data source 102. For example, a transaction server 103 in the transactional system 101 can obtain the RM instance information 104 from the data source 102 by taking advantage of a user callback 110. The transactional system 101 can use different mechanisms for registering the user callback 110, e.g. a static registration mechanism and a dynamic registration mechanism.

The static registration mechanism can be used when a XA server is supported. The XA server is a server that uses the uniform XA interface to control a transaction. For example, in TUXEDO, the static registration can be invoked in the tpopen( ) function after the xa_open( ) function is called. Once the registration is successful, the user callback 110 can be invoked when the transactional server 103 establishes the connection to a database, e.g. an ORACLE® database. Additionally, the user callback 110 can be deregistered in the tpclose( ) function before the xa_close( ) function is called.

Alternatively, the transaction server 103 can dynamically register the user callback 110, e.g. based on a shared library 105 associated with the data source 102. For example, TUXEDO can dynamically register the callback when a user connects to an ORACLE® Database using a non-XA server (e.g. via OCI or Pro*c/c++). TUXEDO can first load ORACLE® OCI library OCI API dynamically and obtain the related OCI environment handle. Then, TUXEDO can register a user callback via an OCIUserCallbackRegister in the OCISessionBegin function.

As shown in FIG. 1, the system can save the obtained instance information 104 in a related context 106 associated with the transactional server 103. Additionally, the transactional server 103 can store the instance information 104 into the different state tables 108 in a shared memory 107 (e.g. the global bulletin board (BB) in TUXEDO). These tables 108 can be synchronized to different nodes, and can be accessed by multiple transactional servers (e.g. servers 111-112) and/or native clients.

FIG. 2 shows an illustration of maintaining various state tables in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 2, a transactional system 200 can store the instance information 220 into different state tables in a shared memory 201.

These state tables can include a resource manager table 211 that stores unique RM/database names, an instance table 212 that stores the RM/database instance names and a service table 213 that stores the RM/database service names. Such information can help other servers to know the instance information associated with the particular server.

Additionally, the transactional system 200 can maintain a server table (ST) 214 in the shared memory 201. The ST 214 can include one or more sever table entries (STEs), each of which can store an index in the instance table 212. For example, each STE can store an instance identifier (ID) if the server is a single-threaded server.

As shown in FIG. 2, the server table 214 can point to the other tables 211-213 in the shared memory 201. Thus, the transactional system 200 can use the server table 214 to obtain the instance information (such as the information on which RM instance that a particular server is currently connected to), and the transactional system 200 may not use the instance information stored in the different state tables 211-213 directly.

FIG. 3 shows an illustration of supporting a server table (ST) in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 3, a server table (ST) 310 in a transactional system 300 can include one or more server table entries (STEs) 311-312, each of which can include one or more instance identifiers (IDs). For example, STE 311 can include instance IDs 321-322, and the STE 312 can include instance IDs 323-324.

In accordance with an embodiment of the invention, each instance ID 321-324 can include various instance information. As shown in FIG. 3, the instance ID 322 can identify an instance name 301, a database name 302, and a service name 303.

For example, the instance ID can be implemented using an integer, e.g. a 32 bit integer that includes three sections: bit 0-11, bit 12-19, and bit 20-31. The first section, bit 0-11, can store the entry index for a RM/database instance name 301. The second section, bit 12-19, can store the entry index for a RM/database name 302. The third section, bit 20-31, can store the entry index for a RM/database service name 303. Additionally, the special value 0xFFFFFFFF can be used for indicating an invalid instance ID.

In accordance with an embodiment of the invention, the transactional system 300 can check the instance information in an instance ID 322 by simply comparing the related bits. Thus, the system can avoid the performance issue due to string comparing, since the string comparing operation is a more expensive than the bit comparing operation.

FIG. 4 shows an illustration of updating the instance information in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 4, a transactional system 401 in a transactional environment 400 can obtain the related instance information 404 from a resource manager (RM) 402, e.g. in a database.

When the user callback 410 is called, the latest instance information 404 can be retrieved from the RM 402 and can be stored into the context 405. Additionally, the transactional system 401 can set a flag 409, which indicates the receiving of the latest instance information 404.

In accordance with an embodiment of the invention a transactional process 403 can be configured with one or more checkpoints. For example, the checkpoints can be triggered before and after the service invocation and after the initialization routine. Also, the check points can be triggered when a connection is established or dropped.

As shown in FIG. 4, at the check point 408, the transactional process 403 can check the flag 409. If the flag 409 is up, the transactional server 403 can update the transaction context 407 based on the retrieved instance information 404, and store the retrieved instance information 404 into the state tables 406 (in a shared memory).

FIG. 5 shows an illustration of supporting a transaction process with various check points in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 5, at step 501, the transactional process starts. Then, the transactional process can proceed with the initialization process at step 502.

The initialization process 502 can include one or more instance check points, e.g. the check point 510. For example, in TUXEDO, when the dynamic registration is used, the check point 510 may be located in the startup routine of a server after the initialization routine (e.g. tpsvrinit). Also, the check point 510 can be triggered when the server tries to establish a connection to RM (e.g. after the xa_open( ) function call is invoked successfully in the tpopen( ) function call).

Additionally, at step 512, the transactional process can retrieve the instance information to enable the instance awareness capability, during the initialization process 502.

Furthermore, at step 503, the transactional process can check whether there is a service request. At step 505, the transactional process can invoke a service dispatcher, e.g. tmsvcdsp( ). As shown in FIG. 5, a check point 506 can be triggered before the service routine 507 is invoked for processing the request message. Additionally, another check point 508 can be triggered after the service routine 507 is completed. At step 509, the transactional process can end the service dispatcher.

Otherwise, if there is no service request and the transactional process is to be shut down, at step 504, the transactional process can start the shutdown process 511. The shutdown process 511 can trigger a check point 513 to clean instance ID information. For example, in TUXEDO, the check point 513 can be triggered when the server is to close a connection to the RM (e.g. before a xa_close( ) function call is invoked in a tpclose( ) function call). Finally, the transactional process ends at step 514.

In accordance with an embodiment of the invention, when the system uses the static registration instead of dynamic registration, the system may behave differently at the check point 512 and the check point 513. The instance information can be retrieved and updated directly without checking the initiator. For example, when the static registration is used, a XA server can use the tpopen( )/tpclose( )function calls to establish/delete the connections to a RM. Also, the tpopen( ) function can be called in the customized tpsvrinit( ) function call, and the tpclose( ) function can be called in the customized tpsvrdone( ) function.

FIG. 6 illustrates an exemplary flow chart for supporting resource manager (RM) instance awareness in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 6, at step 601, a transactional server can receive resource manager (RM) instance information from a resource source that is associated with one or more RM instances, wherein the received instance information allows the transactional server to be aware of which RM instance that the transactional server is currently connected to. Then, at step 602, the system can save the received instance information into one or more tables that are associated with the transactional server. Furthermore, at step 603, the system allows the transactional server to process a global transaction based on the instance information saved in the one or more tables.

Transaction Affinity

FIG. 7 shows an illustration of supporting transaction affinity based on instance awareness in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 7, a transactional environment 700 can support the processing of a global transaction 710 using one or more resource managers (RMs), e.g. a RM instance 702 that is associated with a database 704.

The transactional system 701 can route a request 711 for a database connection to a transactional server 703 (e.g. using the default routing policy). Additionally, the system can assign an affinity context 705 to the transactional server 703. The affinity context 705, which contains information identifying the RM instance 702, can be store in a shared memory (e.g. a global transaction table (GTT) in TUXEDO) and can be propagated using a message. For example, the transactional server 703 can obtain the RM instance name, the RM/database name and the RM/database service name, via the affinity context 705 based on the above instance awareness feature.

Furthermore, a subsequent request 712 can be routed to the transactional server 703 based on the affinity context 705. Additionally, other subsequent requests can also be routed to the transactional server 703, which is connected to the RM instance 702, until the global transaction 710 is completed (or the client context ends).

In accordance with an embodiment of the invention, the transaction affinity ensures that the related database requests 711-712, which are connected to the RM instance 702, can be routed to the same transactional server 703. Thus, the transaction affinity can improve application performance by maximizing the database cluster utilization, since the transaction affinity can improve database performance by increasing the likelihood of cache hits.

In accordance with an embodiment of the invention, the system can apply the transaction affinity routing policy along with other routing polices for performing the global transaction 710. For example, the following routing precedence can be supported in TUXEDO:

1. Transaction precedence routing for domain

2. Client/server affinity routing

3. Transaction affinity routing

4. Load balancing according to service load

FIG. 8 shows an illustration of supporting the transaction affinity routing in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 8, a transactional system can support the processing of a global transaction using a transaction affinity routing policy.

At step 801, the system can receive a transaction request that involves a database connection. Then, at step 802, the system can check whether there are existing affinity contexts.

If no affinity context are involved in the global transaction, the system can perform the load balancing routing, at step 809.

If there are existing affinity contexts involved in the global transaction, the system can apply the transaction affinity routing policy in order to find a server.

At step 803, the system may try to find a server, which is associated with the same instance name, the same database (DB) name and the same service name.

If the system cannot find the server, at step 804, the system may try to find a server, which is associated with the same DB name and the same service name and is in a group that is not involved in the current global transaction.

If the system cannot find the server, at step 805, the system may try to find a server, which is associated with the same DB name and the same instance name.

If the transactional system cannot find the server, at step 806, the system may try to find a server, which is associated with the same DB name.

At step 807, the system may be able to find a server based on the existing affinity contexts. On the other hand, at step 808, the system may not be able to find a server. Then, at step 809, the system can try to find the server according to the load balancing routing.

FIG. 9 shows an illustration of sending a message with an affinity context in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 9, a transactional system 901 can support the transaction processing in a transaction environment 900. Furthermore, a transactional server 903 in a transactional system 901 can support the transaction processing based on a transaction context 907 (e.g. TUXC in TUXEDO).

As shown in FIG. 9, the transactional server 903 can obtain the related affinity context 904 from a shared memory 902 (e.g. GTT in TUXEDO), and can update the transaction context 907 using the related affinity context 914. When the checkpoint 908 is triggered, the system can copy the related affinity context 914 from the transaction context 907 into a message 906 in the message queue 905.

Thus, the transactional system 901 can refer to the related affinity context 914 in the transaction context 907 for service routing, before sending the message 906 to a service.

FIG. 10 shows an illustration of receiving a message with affinity context in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 10, a transactional system 1001 in a transactional environment 1000 can use a message queue 1005 for receiving one or more messages (e.g. a message 1006).

A transactional server 1003 in the transactional system 1001 can continuously dequeue messages (containing requests) after the initial boot. As shown in FIG. 10, the transactional server 1003 can read the message 1006 from the message queue 1005 and processes a service request in the message 1006.

During the processing of the service request, the system can trigger a checkpoint 1008 for copying the affinity context from the message 1006 into the transaction context 1007. Then, the system can use the affinity context 1014 in the transaction context 1007 to update the affinity context 1004 in the shared memory 1002.

Once the request in the message 1006 has been processed, the transactional server 1003 process can read more messages from message queue 1005. Otherwise, the transactional server 1003 can wait on the message queue 1005 until the next request arrives.

FIG. 11 shows an illustration of supporting affinity routing within a client context in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 11, a transactional system 1101 in a transactional environment 1100 can support the transaction processing in a client context 1110 using one or more resource manager (RMs), e.g. a RM instance 1102 that is associated with a database 1104. For example, within the client context 1110, web conversations can be connected and disconnected for a number of times. During each of these connections, the conversation may refer to and/or access the same (or similar) data, such as a shopping cart.

In accordance with an embodiment of the invention, the system can route a request 1111 for a database connection to a transactional server 1103 (e.g. based on the default routing policy). Additionally, the system can assign an affinity context 1105 that indicates the RM 1102 to the transactional server 1103.

Furthermore, one or more subsequent requests within the client context 1110 (e.g. the request 1112) can be routed to the transactional server 1103, based on the affinity context 1105 until the client context 1110 ends or the related transaction completes. Thus, the transactional system 1101 can ensure that the various database operations within the client context 1110 can be directed to the same RM instance 1102.

In accordance with an embodiment of the invention, the transactional system 1101 can receive various load balance advisory events from the database, which hints for an affinity within the client context 1110. For example, in TUXEDO, the load balancing advisory event received from a database can include a parameter, AFFINITY HINT, which is a flag that indicates whether affinity is active or inactive for a particular instance and service combination. The AFFINITY HINT parameter, which is a temporary affinity that lasts for the duration of a web session, can be automatically enabled when load balancing advisory is enabled by setting the goal on the service. Additionally, different instances offering the same service can have different settings for AFFINITY HINT.

In accordance with an embodiment of the invention, the transactional system 1101 may apply the transaction affinity routing policy, instead of the client context based affinity policy, if related database operations are within a transaction. On the other hand, the system can implement the client context based affinity routing policy, based on the default TUXEDO load balance route policy.

FIG. 12 shows an illustration of propagating an affinity context across different domains in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 12, a transactional environment 1200 can support the transaction processing using one or more resource managers (RMs), e.g. a RM instance 1205 that is associated with a database 1215.

In accordance with an embodiment of the invention, the system can route a request 1211 for a database connection to a transactional server 1203 (e.g. using the default routing policy). Additionally, the system can assign an affinity context 1207 that indicates the RM instance 1202 to the transactional server 1203.

Furthermore, a transactional domain in the transactional environment 1200 can propagate affinity context information across domains, if the request 1211 is to be transferred between different domains.

As shown in FIG. 12, the transaction domain 1201 can translate the affinity context 1207 into an affinity key string 1208 before sending the affinity key string 1208 to the remote domain 1202. After receiving the affinity key string 1208, the transaction domain 1202 can translate the affinity key string 1208 into the affinity context 1206 that can be used by a transactional server 1004 in the transaction domain 1202.

Thus, one or more subsequent requests (e.g. the request 1212) can be directed to the RM instance 1202 based on the affinity context 1206.

FIG. 13 shows an illustration of propagating an affinity context to an application server in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 13, a transactional system 1301 in a transactional environment 1300 can support transaction processing using one or more resource manager (RMs), e.g. a RM instance 1305 that is associated with a database 1315.

In accordance with an embodiment of the invention, the system can route a request 1311 for a database connection to a transactional server 1303 (e.g. using the default routing policy). Additionally, the system can assign an affinity context 1307 that indicates the RM instance 1305 to the transactional server 1303.

Furthermore, the transactional system 1301 (e.g. a TUXEDO TDomain) can propagate the affinity context 1307 information to an application server 1302 (e.g. a WEBLOGIC application server) via an affinity key string 1308. For example, the affinity context 1307 can be exchanged between the transactional system 1301 and the application server 1302, via a connector 1304 (e.g. a TUXEDO WTC).

When the transactional system 1301 sends a request to the connector 1304, the transactional system 1301 can translate the affinity context 1307 into the affinity key string 1308. When the connector 1304 receives the request from the transactional system 1301, the connector 1306 can translate the affinity key string 1308 into the transaction context 1306 that can be used by the application server 1302.

When the connector 1304 sends a request to the transactional system 1301, the connector 1304 can obtain affinity context from the transaction context 1306 associated with the application server 1302. When the transactional system 1301 receives the request from the connector 1304, the transactional system 1301 can translate the affinity key string 1308 into the affinity context 1307.

Thus, one or more subsequent requests, e.g. the request 1312, in the application server 1312 can be directed to the RM instance 1302 based on the affinity context 1305.

FIG. 14 illustrates an exemplary flow chart for supporting transaction affinity based on instance awareness in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 14, at step 1401, the system can route a request to a transactional server, wherein the transactional server is connected to a resource manager (RM) instance. Then, at step 1402, the system can assign an affinity context to the transactional server, wherein the affinity context indicates the RM instance that the transactional server is associated with. Furthermore, at step 1403, the system can route one or more subsequent requests that are related to the request to the transactional server based on the affinity context.

Common Transaction Identifier (XID)

FIG. 15 shows an illustration of processing a global transaction in a transactional environment using different transaction identifiers (XI Ds), in accordance with an embodiment of the invention. As shown in FIG. 15, a transactional system 1500 can support the processing of a global transaction 1510 using different resource manager (RM) instances, e.g. RM instances 1504-1505 that connect to a database 1506.

In accordance with an embodiment of the invention, the global transaction 1510 can be associated with a global transaction identifier (GTRID 1520). Within the global transaction 1510, the involved transactional servers in the same group can share one transaction branch, while the transactional servers in different groups may use different transaction branches.

As shown in FIG. 15, the transactional system 1500 may use multiple branches, e.g. branches A-C 1521-1523, for processing the global transaction 1510. Each of the branches A-C 1521-1523 can be associated with a branch qualifier, e.g. BQUALs A-C 1531-1533.

In accordance with an embodiment of the invention, the transactional system 1500 can use different transactional managers (TMs), e.g. TMs A-C 1501-1503, for managing the processing of the global transaction 1510 on different branches A-C 1521-1523.

For example, TM A 1501, which is associated with a transaction identifier (XID) A 1511, can be responsible for managing the processing of the global transaction 1510 on the branch A 1521. TM B 1502, which is associated with a transaction identifier (XID) B 1512, can be responsible for managing the processing of the global transaction 1510 on the branch B 1522. TM C 1503, which is associated with a transaction identifier (XID) C 1513, can be responsible for managing the processing of the global transaction 1510 on branch C 1523.

As shown in FIG. 15, the XI Ds A-C 1511-1513 for the different branches A-C 1521-1523 in a global transaction 1510 can share the same GTRID 1520 (and format ID), and may have different branch qualifiers (i.e. the BQUALs A-C 1531-1533).

In accordance with an embodiment of the invention, the system can invoke a two-phase commit (2PC) process on the global transaction 1510, when more than one groups of transactional servers are involved in the global transaction 1510.

As shown in FIG. 15, the system may use the 2PC model to process the global transaction 1510, even when different participated transactional groups (e.g. TMs A-B 1501-1502) in a global transaction 1510 are in fact associated with same RM instance 1504.

In accordance with an embodiment of the invention, the system can improve the performance of processing the global transaction 1510 via using a common XID when more than one groups run on the same resource manager instance 1504.

FIG. 16 shows an illustration of processing a global transaction in a transactional environment using a common XID, in accordance with an embodiment of the invention. As shown in FIG. 16, a transactional system 1600 can support the processing of a global transaction 1610, which is associated with the GRTID 1620, using the resource manager (RM) instances 1604-1605 that connect to a database 1606.

Furthermore, the transactional system 1600 can use the transactional managers (TMs) A-C 1601-1603 to manage the processing of the global transaction 1610 on different transactional application servers in the different transaction groups (e.g. branches A-C 1621-1623).

In accordance with an embodiment of the invention, each of the resource managers (RMs) 1604-1605 can be uniquely identified, e.g. based on the database name, the server name and the instance name. Based on the database instance awareness capability, a transactional server in the transactional system 1600 is able to know which RM instance that it is currently connecting to.

As shown in FIG. 16, the transactional system 1600 can include a coordinator, such as the TM A 1601. The coordinator TM A 1601 is associated with an instance ID 1641 that identifies the RM instance 1604.

Additionally, the global transaction 1610 can include one or more participating servers (e.g. the participant TMs B-C 1602-1603), which may be located on the local server node where the coordinator 1601 is located or may be located at remote server nodes. Each of the participant TMs B-C 1602-1603 can also be associated with an instance ID (e.g. instance IDs 1642-1643) that identifies the RM instance that it connects to.

In accordance with an embodiment of the invention, when the common XID feature is enabled for processing the global transaction 1610, the XID 1611 for the coordinator TM A 1601 can be shared within the global transaction 1610 (i.e. the XID 1611 is used as a common XID). Thus, a transactional application that involves multiple groups and runs on a clustered database, such as an ORACLE® database, can improve the transaction performance by taking advantage of the database instance awareness.

As shown in FIG. 16, the coordinator TM A 1601 for the global transaction 1610 can propagate various types of information, such as the XID 1611 and the instance ID 1641, to different participant TMs B-C 1602-1603, within the life cycle of the global transaction 1610.

Furthermore, each of the participant TMs B-C 1602-1603 can determine whether it shares the same RM with the coordinator TM A 1601 by comparing the received instance ID 1641 with its own instance ID. If the instance IDs are the same, the participant TMs B-C 1602-1603, can mark itself as a common-XID server (or group).

For example, the system may find a match on the branch B 1622, since the TM B 1602 shares the same RM instance 1604 with the TM A 1601. Thus, the TM B 1602 can use the common XID 1611 instead of its own XID for supporting transaction processing. Then, the TM B 1602 can notify the coordinator TM A 1601 that it uses the common XID 1611. In such a case, when the coordinator TM A 1601 operates to commit or rollback the global transaction 1610, the system can ignore the branch B 1622, since it uses the common XID 1611 (which is based on the BQUAL A 1631).

On the other hand, the system may not find a match on the branch C 1623 (which uses the XID C 1613 and BQUAL C 1633), since the TM C 1603 is associated with a different RM 1605 (with an instance ID 1643). Optionally, the TM C 1603 can notify the coordinator TM A 1601 that it does not use the common XID 1611. Then, the system can process the transaction branch C 1623 according to the two-phase commit (2PC) processing model.

In accordance with an embodiment of the invention, the different common-XID groups, i.e., the groups of transactional servers that are associated with the same RM instance 1604 with the coordinator 1601, can access the RM instance 1604 via the common XID 1611.

Furthermore, when the commit request is invoked, the coordinator 1601 may not send any messages to the local common-XID groups. The system can change the state of each local common-XID group to read-only at once. Also, the remote common-XID groups can receive the prepare request from the coordinator 1601 and may change its state to read-only without any real database operation. Thus, the system may only need to prepare/commit one of these groups (e.g. the coordinator's group).

Additionally, the system can change a common-XID group to become a non-common-XID group, if the instance ID changes. For example, if branch B 1622 changes to use another RM instance, then the system can invoke a two-phase commit (2PC) process instead (e.g. based on the BQUAL B 1632).

FIG. 17 shows an illustration of supporting a one-phase commit (1PC) processing model based on database instance awareness in a transactional environment, in accordance with an embodiment of the invention. As shown in FIG. 17, a transactional system 1700 can support the processing of a global transaction 1710, which is associated with the GRTID 1720, using the resource manager (RM) instance 1704 that connects to a database 1706.

The transactional system 1701 can include a plurality of transactional managers (TMs) A-C 1701-1703, which are used to manage the processing of the global transaction 1710 in different transactional groups (i.e. branches A-C 1721-1723).

Furthermore, the TMs A-C 1701-1703 can manage the processing of the global transaction 1710 based on a single resource manager (RM) instance 1704. Each of the TMs A-C 1701-1703 can maintain an instance identifier (ID), e.g. the instance IDs 1741-1743.

As shown in FIG. 17, the transactional system 1700 can include a coordinator (e.g. the TM A 1701). The coordinator TM A 1701 for the global transaction 1710 can propagate various types of information, such as the common XID 1711 and the instance ID 1741, to the different participating transactional servers (e.g. the TMs B-C 1702-1703), within the life cycle of the global transaction 1710.

In accordance with an embodiment of the invention, based on the transaction affinity capability, the system can route all related requests in the global transaction 1710 to the same RM instance 1704. Furthermore, based on the instance awareness capability, the TMs A-C 1701-1703 are able to know that only a single RM instance 1704 is used in the global transaction 1710, since the different instance IDs 1741-1743 all identify the same RM instance 1704. Thus, the coordinator TM A 1701 can uses the common XID 1711 (e.g. its own XID that is based on the BQUAL A 1731 and the GRTID 1720) for coordinating the global transaction 1710 in the transactional environment 1700.

As shown in FIG. 17, the coordinator TM A 1701 may not send any “prepare/commit” request to other groups at the commit stage, since they are all common-XID groups. Furthermore, the system can take advantage of the one-phase commit (1PC) processing model.

In accordance with an embodiment of the invention, the read-only one-phase commit optimization can significantly improve the system performance. It can perform 1 PC processing on the reserved group if all other groups returns read-only. The performance can be improved when all branches of the global transaction are tightly coupled, e.g. in the same instance or the same RAC.

For example, a transactional environment 1700 may have N (where N>1) participated groups for processing the global transaction 1710. Among them, M (where M<N) participated groups may have the same instance ID as that of the coordinator.

Using the two-phase commit (2PC) processing model with no database instance awareness, the system may perform N prepare operations and one commit operation on a database (e.g. as shown in FIG. 15). Also, the system may need to write a transaction log.

Alternatively, based on the database instance awareness (e.g. as shown in FIG. 16), the system may need to perform N-M prepare operations and one commit operation on a database (with M prepare operations reduced).

Furthermore, when M=N−1, which indicates that all other participated groups in the global transaction can share the same transaction branch with the coordinator. Then, there is only one branch in processing the global transaction. The system may only need to perform one commit operation, with N (or M+1) prepare operations reduced. Also, the system may not need to write a transaction log.

FIG. 18 shows an illustration of processing a global transaction based on database instance awareness in a transactional middleware environment, in accordance with an embodiment of the invention. As shown in FIG. 18, a transactional system, e.g. an ORACLE® TUXEDO system, can support the processing of a global transaction using multiple transactional groups, e.g. TUXEDO groups A-B 1802-1803.

Furthermore, each of the TUXEDO groups A-B 1802-1803 can have a set of transaction manager servers (TMS). For example, group A 1802 includes a server A 1804 and a TMS A 1806, which can act as a coordinator. Additionally, group A 1802 can include a shared memory, e.g. a TUXEDO bulletin board (BB) A 1808. Furthermore, group B 1803 includes a server B 1805 and a TMS B 1807, and a shared memory, e.g. a TUXEDO BB B 1809.

In the example as shown in FIG. 18, at step 1811, the client 1801 can access a service (e.g. Service A) on the server A 1804 by invoking a function call, tpcall(Service A). Then, at step 1812, the server A 1804 can create related global transaction table entries (GTTEs) in the BB A 1808.

Additionally, at step 1813, the client 1801 can access another service (e.g. Service B) on the server B 1805 by invoking a function call, tpcall(Service B). At step 1814, the server A 1804 can add the related information about group A 1802 into the BB B 1809. Also, at step 1815, the client 1801 can add the related information about group B 1803 into the BB A 1808.

Furthermore, at step 1816, the client 1801 can request to commit the transaction by invoking a function call, tpcommit( ). TUXEDO can invoke 1PC directly on the global transaction if all groups involved in the global transaction run on the same RM instance. At step 1817, the coordinator TMS A 1806 can proceed to commit the global transaction.

If the 1PC call succeeds, at step 1818, the coordinator TMS A 1806 can remove the GTTE in the local node. Then, at step 1819, the coordinator 1806 can notify the Group B 1803, a remote common-XID group, to forget its branch. Finally, at step 1820, the TMS B 1807 can update the BB B 1809.

FIG. 19 shows an illustration of processing a global transaction across multiple domains in a transactional environment using a common XID, in accordance with an embodiment of the invention. As shown in FIG. 19, a transactional system 1900 can support the processing of a global transaction 1910 across multiple domains, e.g. domains A-B 1951-1952, based on the resource manager (RM) instance 1904 that connect to a database 1906.

Additionally, the different branches A-B 1921-1923 can share a GTRID 1920 in the transactional system 1900. The coordinator TM A 1901 in domain A 1951 can propagate the transaction identifier (XID) 1911 and the instance information based on the instance ID 1941 to the remote domain B 1952.

In accordance with an embodiment of the invention, the instance ID 1941, which is unique within the domain A 1951, may be different in different domains (e.g. the domain B 1952), due to the different server booting up sequences.

As shown in FIG. 19, instead of propagating the instance ID 1941 crossing domains directly, the coordinator TM A 1901 can convert the instance ID 1941 into a formatted string 1908, before being propagated across domains. For example, the formatted string 1908 can include the database name, the server name and the instance name.

Additionally, a domain gateway server 1905 can be used for supporting the communication between the local domain A 1951 and the remote domain B 1952. The outbound interface of the domain gateway server 1905 can map the instance information from the instance ID 1941 into the formatted string 1908. The inbound of the domain gateway server 1905 can map the instance information from the formatted string 1908 into the instance ID 1941.

For example, in TUXEDO, the user may divide the TUXEDO groups into different domains for business reasons. A gateway server, such as the GWTDOMAIN server, can be used to support the communication between the different domains. Furthermore, the GWTDOMAIN server in the coordinator domain can function as a proxy. Additionally, the GWTDOMAIN server can be configured to use a common XID, when all involved servers in the other domains through the GWTDOMAIN server are set to use the common XID.

In accordance with an embodiment of the invention, the remote domain B 1952 can store the common XID 1911 as an imported XID 1913. As shown in FIG. 19, the branch B 1923 (i.e. a common-XID group) can use the imported XID 1943 for accessing the database, if the imported XID 1913 exists and the BQUAL A 1931 associate with the imported XID 1913 is valid.

In accordance with an embodiment of the invention, based on database instance awareness, the system can also take advantage of one-phase commit (1PC) model in processing an inter-domain transaction, when the inter-domain transaction involves a single RM instance.

FIG. 20 illustrates an exemplary flow chart for processing a global transaction in a transactional environment using a common XID, in accordance with an embodiment of the invention. As shown in FIG. 20, at step 2001, a coordinator for the global transaction can propagate a common transaction identifier and information for a resource manager instance to one or more participants of the global transaction in the transactional environment. Then, at step 2002, the system allows said one or more participants, which share resource manager instance with the coordinator, to use the common transaction identifier. Furthermore, at step 2003, the coordinator can process the global transaction for said one or more participants that share the resource manager instance using one transaction branch.

Many features of the present invention can be performed in, using, or with the assistance of hardware, software, firmware, or combinations thereof. Consequently, features of the present invention may be implemented using a processing system (e.g., including one or more processors).

Features of the present invention can be implemented in, using, or with the assistance of a computer program product which is a storage medium (media) or computer readable medium (media) having instructions stored thereon/in which can be used to program a processing system to perform any of the features presented herein. The storage medium can include, but is not limited to, any type of disk including floppy disks, optical discs, DVD, CD-ROMs, microdrive, and magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic or optical cards, nanosystems (including molecular memory ICs), or any type of media or device suitable for storing instructions and/or data.

Stored on any one of the machine readable medium (media), features of the present invention can be incorporated in software and/or firmware for controlling the hardware of a processing system, and for enabling a processing system to interact with other mechanism utilizing the results of the present invention. Such software or firmware may include, but is not limited to, application code, device drivers, operating systems and execution environments/containers.

Features of the invention may also be implemented in hardware using, for example, hardware components such as application specific integrated circuits (ASICs). Implementation of the hardware state machine so as to perform the functions described herein will be apparent to persons skilled in the relevant art.

Additionally, the present invention may be conveniently implemented using one or more conventional general purpose or specialized digital computer, computing device, machine, or microprocessor, including one or more processors, memory and/or computer readable storage media programmed according to the teachings of the present disclosure. Appropriate software coding can readily be prepared by skilled programmers based on the teachings of the present disclosure, as will be apparent to those skilled in the software art.

While various embodiments of the present invention have been described above, it should be understood that they have been presented by way of example, and not limitation. It will be apparent to persons skilled in the relevant art that various changes in form and detail can be made therein without departing from the spirit and scope of the invention.

The present invention has been described above with the aid of functional building blocks illustrating the performance of specified functions and relationships thereof. The boundaries of these functional building blocks have often been arbitrarily defined herein for the convenience of the description. Alternate boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Any such alternate boundaries are thus within the scope and spirit of the invention.

The foregoing description of the present invention has been provided for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. The breadth and scope of the present invention should not be limited by any of the above-described exemplary embodiments. Many modifications and variations will be apparent to the practitioner skilled in the art. The modifications and variations include any relevant combination of the disclosed features. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications that are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the following claims and their equivalents.

Claims

1. A method for supporting transaction processing in a transactional system comprising:

a plurality of transactional servers including a first group of transactional servers including a first transaction manger and a second group of transactional servers including a second transaction manager, and a plurality of resource manager instances for managing access to a database, the method comprising:

receiving a global transaction from said plurality of transactional servers directed via said first transaction manager and said second transaction to a shared resource manager instance of said plurality of resource manager instances;

using the first transaction manager in the first group of transactional servers as the coordinator for said global transaction;

propagating, via the coordinator for said global transaction, a common transaction identifier and information for said resource manager instance to both of said first transaction manager and second transaction manager participating in said global transaction in the transactional environment;

allowing said first transaction manager and second transaction manager which share said shared resource manager instance, to use the common transaction identifier; and

processing the global transaction, via the coordinator and shared resource manager instance using one transaction branch.

2. The method of claim 1, further comprising:

using a transaction identifier assigned to the global transaction by the first transaction manager in first group of transactional servers as the common transaction identifier.

3. The method of claim 1, wherein

said database comprises a clustered database.

4. The method of claim 1, further comprising:

committing or rolling back the global transaction, via the coordinator, for said first transaction manager and second transaction manager.

5. The method of claim 1, further comprising:

using another transaction branch to process the global transaction for a third transaction manager participating in the global transaction, wherein said third transaction manager utilizes a different resource manager instance of said plurality of resource manager instances than said shared resource manager instance.

6. The method of claim 1, further comprising:

invoking a one-phase commit operation directly on the shared resource manager instance, when all transaction managers participating in the global transaction share the shared resource manager instance.

7. The method of claim 1, further comprising:

sending a forget request to a transaction manager on a remote node in order to release remote transactional resources.

8. The method of claim 1, further comprising:

allowing the first and second transaction managers to be in different domains.

9. The method of claim 1, wherein the first transaction manager is in a remote domain from the second transaction manager, and wherein the method further comprises:

receiving the common transaction identifier and the identification information for the resource manager instance at the second transaction manager from the remote domain via a gateway.

10. A system for supporting transaction processing in a transactional environment, comprising:

a plurality of networked computer systems each comprising one or more microprocessor and memory;

a plurality of transactional servers operating on said plurality of networked computer systems, wherein said plurality of transactional servers include a first group of transactional servers including a first transaction manger, and a second group of transactional servers including a second transaction manager;

a database;

a plurality of resource manager instances for managing access to the database;

wherein said plurality of transactional servers are configured to generate a global transaction directed via said first transaction manager and said second transaction to a shared resource manager instance of said first plurality of resource manager instances and second plurality of resource manager instances;

wherein the first transaction manager functions as the coordinator for the global transaction;

wherein the coordinator for said global transaction is configured to propagate a common transaction identifier and information for said resource manager instance to both of said first transaction manager and second transaction manager participating in said global transaction in the transactional environment;

wherein said first transaction manager and second transaction manager which share said shared resource manager instance are configured to use the common transaction identifier to coordinate the global transaction; and

wherein the coordinator is configured to process the global transaction, via the coordinator and shared resource manager instance to the database, using one transaction branch.

11. The system according to, claim 10, wherein:

a transaction identifier assigned to the global transaction by the first transaction manager in the first group of transactional server is used as the common transaction identifier.

12. The system according to claim 10, wherein:

the database is a clustered database.

13. The system according to claim 10, wherein:

the coordinator is configured to commit or roll back the global transaction for both of said first transaction manager and second transaction manager.

14. The system according to claim 10, further comprising:

a third transaction manager participating in said global transaction, wherein said third transaction manager utilizes a different resource manager instance of said plurality of resource manager instances than said shared resource manager instance; and

wherein another transaction branch is used to process the global transaction for said third transaction manager via said different resource manager.

15. The system according to claim 10, wherein:

an one-phase commit operation is invoked directly on the resource manager instance, when all transaction managers participating in the global transaction share the shared resource manager instance.

16. The system according to claim 10, wherein:

the coordinator is further configured to send a forget request to a remote transaction manager on a remote node in order to release remote transactional resources.

17. The system according to claim 10, wherein:

the first transaction manager and the second transaction manager are in different domains.

18. A non-transitory machine readable storage medium having instructions stored thereon for supporting transaction processing in a transactional system comprising:

a plurality of transactional servers including a first group of transactional servers including a first transaction manger and a second group of transactional servers including a second transaction manager, and a plurality of resource manager instances for managing access to a database, which instructions, when, executed, cause the transactional system to perform steps comprising:

receiving a global transaction from said plurality of transactional servers directed via said first transaction manager and said second transaction to a shared resource manager instance of said first plurality of resource manager instances and second plurality of resource manager instances;

using the first transaction manager in the first group of transactional servers as the coordinator for said global transaction;

propagating, via the coordinator for said global transaction, a common transaction identifier and information for said resource manager instance to both of said first transaction manager and second transaction manager participating in said global transaction in the transactional environment;

allowing said first transaction manager and second transaction manager which share said shared resource manager instance, to use the common transaction identifier; and

processing the global transaction, via the coordinator and shared resource manager instance using one transaction branch.